Gerçek Zamanlı Yüz Tanıma Tabanlı Personel Kontrol ve Takip Sistemi Tasarımı

Abstract

Image processing and computer vision have started to spread rapidly in recent years. Computer vision studies are carried out in many fields such as agriculture, medicine, education, health and security and are used in our daily life. Especially face and object recognition process comes up with many different applications day by day. Thanks to these face recognition applications developed, databases are created by collecting facial information from users. These applications and databases will facilitate the control and monitoring of many social areas such as schools, workplaces, sites, shopping centers, stadiums. In this article, it is proposed to design a face recognition based personnel control and tracking system in order to track the personnel entrances and exits quickly, effectively and accurately. Controlling the entrances and exits of employees in various workplaces is generally done by card system or by keeping a manual record. In such systems, human-related problems such as printing cards or typing incomplete names occur instead of someone else. In addition, there are serious problems in tracking the number of personnel entrances and exits in companies. Especially after recent epidemic diseases, it has turned out to be a very troublesome situation in terms of hygiene and health conditions for everyone to scan cards or to have fingerprints in the same place. In this study, face recognition based personnel control and tracking system design was developed. In the developed system, cameras are placed on the inputs and outputs and the images of the personnel are taken. After determining the facial regions from the images taken, it is determined to whom the face belongs by using eigenface, fisherface and local binary pattern histogram methods. The entry or exit information of the person found is displayed on the screen and this information is stored in the personnel database. Thanks to the face recognition system, personnel entry-exit times and overtime information are easily tracked automatically.